100. While detailed examination of the area was not possible after the fire, there was no evidence produced at the Inquiry which would suggest that it was initiated by any means other than the spontaneous heating of coal in the roof of the Loader Mine. It must, of course, be conjectural how sufficient air got behind the lining to initiate and sustain a heating. W. Rowell, the deputy director (operations), suggested that there must have been intake to intake leakage but this view is only valid if one accepts that the seals into the sides of the Loader Mine or those along the sides at floor level were defective. There was no evidence to this effect. It is difficult to be more precise than to say that the heating occurred in the roof coal somewhere in that reach of the Loader Mine which had been lined with polyurethane foam. But it is, I think, probable that the source of air leakage which caused this was in the direct connection with the No. 3 Heading between the Loco Level and Dysart Main Return between which there was a difference in water gauge of 1½ inches. In considering just how the fire developed two facts are, I think, of outstanding importance. First, there was in a very short period of time the production of large volumes of thick smoke and, second, that such carbon monoxide as was produced was diluted to less than lethal concentration. In this latter connection it is important to note that, although all the six men whose bodies were recovered showed some signs of carbon monoxide poisoning, there was no marked degree of this in any of the large number of people who escaped.
101. It is known that concrete slabs can become cracked or broken when subject to high temperatures and that heated coal loses its inherent strength as combustion progresses. It is probable that the falling of burning material observed by Stirling was continued as a larger fall developed when the burning coal, bereft of its inherent strength, became a dead weight on the concrete slabs and the foam lining. And it may be that the burning of this larger fall accounts for the sudden issue of smoke which apparently occurred in the minute or two which it took Stirling to hurry round to the other end of the Loader Mine. However, the experiment I witnessed at Buxton showed that the burning seen by Stirling would have sufficed to ignite the foam and so bring about a polyurethane fire, which would have produced the same sudden issue of heavy smoke. What happened after this is again a matter for conjecture but the absence of high concentrations of carbon monoxide in the diluted products suggests that the rate of ventilation through the Loader Mine was reduced by the fire. The evidence of the Buxton experiments leads one to think that the fire fluctuated in intensity both up and downwind in the Loader Mine. Here again, the burning of the P.V.C. belting would, I think, be involved in much the same way. That is to say it did not burn uniformly, but in patches as it became subject to higher temperatures produced by burning foam. This, and a reduced rate of ventilation, would account for the continued production of large volumes of smoke and fumes containing relatively small quantities of carbon monoxide, but dangerous quantities of other toxic gases. Once sufficient heat had been generated by the initial fire there were of course other materials, such as coal and untreated timber, available to it.
102. Both Sir Andrew Bryan and Mr. Hill drew attention to the virtual absence of warning of the initial heating. Laboratory experiments, however, have shown that the foam is such a good sealant in that it not only keeps out air but effectively masks any temperature changes, noxious gas emissions and the smell associated with heatings. Thus all the signs by which colliery personnel might have detected a heating were absent. This would explain why Dryburgh, some time after midnight, noticed nothing untoward. Moreover when Izatt reported to Stirling that he heard rumbling noises — these could well have been strata movements or even the cracking of concrete lining blocks — Stirling made an inspection of the Loader Mine; this took place less than an hour before the fire broke out but he noticed nothing unusual.
103. That a fire-retardant coating had not been applied to the foam in the Loader Mine was not disputed. The management explained that application of the coating had to be deferred while another fire in the Dysart Dip No. 1 Section was dealt with. This may well be so but I feel sure that had they recognised just how inflammable polyurethane foam is — and I take their evidence as meaning that until they saw the Buxton experiments they did not — they would have proceeded with the placing of the coating notwithstanding the demand for men to deal with the other fire. This is not to say that a retardant coating would necessarily have prevented the fire. Indeed, I recognise the possibility that burning coal breaking through the foam lining could well have left a hole with a ragged edge and the then exposed foam would have been susceptible to ignition either by coal burning on the floor or by fire in the cavity.
104. I consider that polyurethane foam as at present constituted is an unacceptable risk and that its use in, and near the entrances to, all mines should be discontinued. I understand that almost a mile of roadways have been treated with polyurethane foam but, notwithstanding the difficulties involved, I am sure that the fire hazard warrants its complete removal or sealing off as soon as practicable.